The orbitals is just a basis function!!! And a s, p or d orbitals are solutions just for the Hydrogen atom. Which means that basing our solutions on it is an approximation anyway!
Besides ... although important to ann accurate description of the total energy, the polarization is a seccondary effect in terms of electronic density. So no reason to care about the information you are losing because the formn of the polarization function. In the density of states, these residual d-orbitals charges will follow the density of the 1st order d terms. Anyway ... it is just my opinion Cheers! NH On Wed, Jan 26, 2011 at 8:51 PM, Weslley Souza Patrocinio < [email protected]> wrote: > Okay. Siesta takes into account the l+1 orbital to consider the > deformations induced by bonding. In my case, the last occupied orbital in Si > has l=1, so the polarization orbital has l=2 (d orbital). However, if the d > "perturbing" state are perturbing the 3p one, when I choose n=3 and l=1 to > view the density of states related to the 3p orbital, how much information > i'm losing due the occupation of d orbitals ? In this case I must to > consider the "d" PDOS as a contribution to the 3p orbital ? > > I read the manual and the article that have this formalism, but this point > is not clear for me. > > Thanks for your help. > > My bests, > > Weslley. >
